Catalytically active monoclonal antibodies (Abs) can be used for selective prodrug activation and chemical transformations. Monoclonal Abs with aldolase activity have emerged as highly efficient catalysts for a number of chemical transformations, particularly aldol and retro-aldol reactions. The retro-aldolase activity of Abs, such as 38C2 and 93F3, have allowed researchers to design, synthesize, and evaluate prodrugs of various chemotherapeutic agents that can be activated by retro-aldol reactions. (Construction of 38C2 was described in WO 97/21803, herein incorporated by reference). 38C2 contains an antibody combining site that catalyzes the aldol addition reaction between an aliphatic donor and an aldehyde acceptor. In a syngeneic mouse model of neuroblastoma, systemic administration of an etoposide prodrug and intra-tumor injection of 38C2 inhibited tumor growth.
One drawback in the use of catalytic Abs is that they lack a device to target the catalytic Ab to the malignant cells. Previous studies demonstrated that in an antibody-directed enzyme prodrug therapy (ADEPT) or antibody-directed abzyme prodrug therapy (ADAPT) approach, enzymes or catalytic antibodies can be directed to tumor cells by chemical conjugation or recombinant fusion to targeting antibodies. However, a more efficient alternative would be using the catalytic antibody fused to a targeting peptide located outside the antibody combining site, thereby leaving the active site available for the prodrug activation. For example, the fusion of Ab 38C2 to an integrin αvβ3-binding peptide would selectively localize the antibody to the tumor and/or the tumor vasculature and trigger prodrug activation at that site. The potential therapy of this approach is supported by preclinical and phase III clinical data suggesting that peptides can be converted into viable drugs through fusion to antibody Fc regions.
The development of bispecific or multi-specific antibodies that target two or more cancer targets simultaneously and or activate prodrugs offers a novel and promising solution to attacking cancer and other diseases. Such antibodies are exemplified in FIG. 1 of the present application. Studies of bispecific antibodies (BsAb) that simultaneously target two tumor-associated antigens (e.g. growth factor receptors) for down-regulation of multiple cell proliferation/survival pathways has provided support for this approach. Traditionally, bispecific antibodies have been prepared by chemically linking two different monoclonal antibodies or by fusing two hybridoma cell lines to produce a hybrid-hybridoma. Dual-specific, tetravalent IgG-like molecules, or dual-variable-domain immunoglobins, have been engineered from two monoclonal antibodies. These dual-variable-domain immunoglobins are capable of binding both antigens in the presence of serum. However, these approaches present challenges with respect to manufacturing, yield and purity.
A variety of recombinant methods have been developed for efficient production of small BsAb fragments such as diabody, minibody, and Fab-scFv fusion proteins. These BsAb fragments may possess some advantages over the full-length IgG-like molecules for certain clinical applications, such as for tumor radio-imaging and targeting, because of better tissue penetration and faster clearance from the circulation. On the other hand, IgG-like BsAb may prove to be preferred over smaller BsAb fragments for other in vivo applications, specifically for oncology indications, by providing the Fc domain that confers long serum half-life and supports secondary immune function, such as antibody-dependent cellular cytotoxicity and complement-mediated cytotoxicity. Unlike their fragment counterparts, engineering and production of recombinant IgG-like BsAb has been, however, rather technically challenging due to their large size (˜150-200 kDa) and structural complexity. Success in the field, as judged by successful application in animal models, has been very limited. Recently, with the examination of a variety of constructs, the efficient expression of Fc domain containing BsAb molecules in mammalian cells has made some strides.
Another approach that has been used to target antibodies is through the use of peptibodies. Peptibodies are essentially peptide fusions with antibody Fc regions. Given the success of studies using random peptide libraries to find high-affinity peptide ligands for a wide variety of targets, fusion of such peptides to antibody Fc regions provides a means of making peptides into therapeutic candidates by increasing their circulatory half-life and activity through increased valency.
Protein interactions with other molecules is basic to biochemistry. Protein interactions include receptor-ligand interactions, antibody-antigen interactions, cell-cell contact and pathogen interactions with target tissues. Protein interactions can involve contact with other proteins, with carbohydrates, oligosaccharides, lipids, metal ions and the like materials. The basic unit of protein interaction is the region of the protein involved in contact and recognition, and is referred to as the binding site or target site.
Peptides derived from phage display libraries typically retain their binding characteristics when linked to other molecules. Specific peptides of this type can be treated as modular specificity blocks or molecular recognition domains (MRDs) that can be combined to create a single protein with binding specificities for several defined targets.
An example of a such a defined target site is integrin. Integrins are a family of transmembrane cell adhesion receptors that are composed of α and β subunits and mediate cell attachment to proteins within the extracellular matrix. At present, eighteen α and eight β subunits are known; these form 24 different αβ heterodimers with different specificities for various ECM cell-adhesive proteins. Ligands for various integrins include fibronectin, collagen, laminin, von Willebrand factor, osteopontin, thrombospondin, and vitronectin, which are all components of the ECM. Certain integrins can also bind to soluble ligands such as fibrinogen or to other adhesion molecules on adjacent cells. Integrins are known to exist in distinct activation states that exhibit different affinities for ligand. Recognition of soluble ligands by integrins strictly depends on specific changes in receptor conformation. This provides a molecular switch that controls the ability of cells to aggregate in an integrin dependent manner and to arrest under the dynamic flow conditions of the vasculature. This mechanism is well established for leukocytes and platelets that circulate within the blood stream in a resting state while expressing non-activated integrins. Upon stimulation through proinflammatory or prothrombotic agonists, these cell types promptly respond with a number of molecular changes including the switch of key integrins, β2 integrins for leucocytes and αvβ3 for platelets, from “resting” to “activated” conformations. This enables these cell types to arrest within the vasculature, promoting cell cohesion and leading to thrombus formation.
It has been demonstrated that a metastatic subset of human breast cancer cells expresses integrin αvβ3 in a constitutively activated form. This aberrant expression of αvβ3 plays a role in metastasis of breast cancer as well as prostate cancer, melanoma, and neuroblastic tumors. The activated receptor strongly promotes cancer cell migration and enables the cells to arrest under blood flow conditions. In this way, activation of αvβ3 endows metastatic cells with key properties likely to be critical for successful dissemination and colonization of target organs. Tumor cells that have successfully entered a target organ may further utilize αvβ3 to thrive in the new environment, as αvβ3 matrix interactions can promote cell survival and proliferation. For example, αvβ3 binding to osteopontin promotes malignancy and elevated levels of osteopontin correlate with a poor prognosis in breast cancer.
For these reasons, and for its established role in angiogenesis, the αvβ3 integrin is one of the most widely studied integrins. Antagonists of this molecule have significant potential for use in targeted drug delivery. One approach that has been used to target αvβ3 integrin uses the high binding specificity to αvβ3 of peptides containing the Arg-Gly-Asp (RGD) sequence. This tripeptide, naturally present in extracellular matrix proteins, is the primary binding site of the αvβ3 integrin. However, RGD based reporter probes are problematic due to fast blood clearance, high kidney and liver uptake and fast tumor washout. Chemical modification of cyclised RGD peptides has been shown to increase their stability and valency. These modified peptides are then coupled to radio-isotypes and used either for tumor imaging or to inhibit tumor growth.
Integrin αvβ3 is one of the most well characterized integrin heterodimers and is one of several heterodimers that have been implicated in tumor-induced angiogenesis. While sparingly expressed in mature blood vessels, αvβ3 is significantly up-regulated during angiogenesis in vivo. The expression of αvβ3 correlates with aggressiveness of disease in breast and cervical cancer as well as in malignant melanoma. Recent studies suggest that αvβ3 may be useful as a diagnostic or prognostic indicator for some tumors. Integrin αvβ3 is particularly attractive as a therapeutic target due to its relatively limited cellular distribution. It is not generally expressed on epithelial cells, and minimally expressed on other cell types. Furthermore, αvβ3 antagonists, including both cyclic RGD peptides and monoclonal antibodies, significantly inhibit cytokine-induced angiogenesis and the growth of solid tumor on the chick chorioallantoic membrane.
Another integrin heterodimer, αvβ5, is more widely expressed on malignant tumor cells and is likely involved in VEGF-mediated angiogenesis. It has been shown that αvβ3 and αvβ5 promote angiogenesis via distinct pathways: αvβ3 through bFGF and TNF-a, and αvβ5 through VEGF and TGF-α. It has also been shown that inhibition of Src kinase can block VEGF-induced, but not FGF2-induced, angiogenesis. These results strongly imply that FGF2 and VEGF activate different angiogenic pathways that require αvβ3 and αvβ5, respectively.
Integrins have also been implicated in tumor metastasis. Metastasis is the primary cause of morbidity and mortality in cancer. Malignant progression of melanoma, glioma, ovarian, and breast cancer have all been strongly linked with the expression of the integrin αvβ3 and in some cases with αvβ5. More recently, it has been shown that activation of integrin αvβ3 plays a significant role in metastasis in human breast cancer. A very strong correlation between expression of αvβ3 and breast cancer metastasis has been noted where normal breast epithelia are αvβ3 negative and approximately 50% of invasive lobular carcinomas and nearly all bone metastases in breast cancer express αvβ3. Antagonism of αvβ3 with a cyclic peptide has been shown to synergize with radioimmunotherapy in studies involving breast cancer xenografts.
Angiogenesis, the formation of new blood vessels from existing ones, is essential to may physiological and pathological processes. Normally, angiogenesis is tightly regulated by pro- and anti-angiogenic factors, but in the case of diseases such as cancer, ocular neovascular disease, arthritis and psoriasis, the process can go awry. The association of angiogenesis with disease has made the discovery of anti-angiogenic compound attractive. The most promising phage derived anti-angiogenic peptide described to date, developed by Amgen, neutralizes the angiogenic cytokine Ang2.
While the VEGFs and their receptors have been among the most extensively targeted molecules in the angiogenesis field, preclinical efforts targeting the more recently discovered angiopoietin-Tie2 pathway are underway. Both protein families involve ligand receptor interactions, and both include members whose functions are largely restricted postnatally to endothelial cells and some hematopoietic stem cell lineages. Tie-2 is a receptor tyrosine kinase with four known ligands, angiopoietin-1 (Ang1) through angiopoietin-4 (Ang4), the best studied being Ang1 and Ang2. Ang1 stimulates phosphorylation of Tie2 and the Ang2 interaction with Tie2 has been shown to both antagonize and agonize Tie2 receptor phosphorylation. Elevated Ang2 expression at sites of normal and pathological postnatal angiogenesis circumstantially implies a proangiogenic role for Ang2. Vessel-selective Ang2 induction associated with angiogenesis has been demonstrated in diseases including cancer. In patients with colon carcinoma, Ang2 is expressed ubiquitously in tumor epithelium, whereas expression of Ang1 in tumor epithelium was shown to be rare. The net gain of Ang2 activity has been suggested to be an initiating factor for tumor angiogenesis.
Other fusion proteins directed towards cellular receptors are under clinical evaluation. Herceptin (Trastuzumab), developed by Genentech, is a recombinant humanized monoclonal antibody directed against the extracellular domain of the human epidermal tyrosine kinase receptor 2 (HER2 or ErbB2). The HER2 gene is overexpressed in 25% of invasive breast cancers, and is associated with poor prognosis and altered sensitivity to chemotherapeutic agents. Herceptin blocks the proliferation of ErbB2-overexpressing breast cancers, and is currently the only ErbB2 targeted antibody therapy approved by the FDA for the treatment of ErbB2 over-expressing metastatic breast cancer (MBC). In normal adult cells, few ErbB2 molecules exist at the cell surface ˜20,000 per cell, so few heterodimers are formed and growth signals are relatively weak and controllable. When ErbB2 is overexpressed, ˜500,000 per cell, multiple ErbB2 heterodimers are formed and cell signaling is stronger, resulting in enhanced responsiveness to growth factors and malignant growth. This explains why ErbB2 overexpression is an indicator of poor prognosis in breast tumors and may be predictive of response to treatment.
ErbB2 is a promising and validated target for breast cancer, where it is found both in primary tumor and metastatic sites. Herceptin induces rapid removal of ErbB2 from the cell surface, thereby reducing its availability to heterodimerize and promote growth. Mechanisms of action of Herceptin observed in experimental in vitro and in vivo models include inhibition of proteolysis of ErbB2's extracellular domain, disruption of downstream signaling pathways such as phosphatidylinositiol 3-kinase (PI3K) and mitogen-activated protein kinase (MAPK) cascades, GI cell-cycle arrest, inhibition of DNA repair, suppression of angiogenesis and induction of antibody dependent cellular cytotoxicity (ADCC). The majority of patients with metastatic breast cancer who initially respond to Herceptin, however, demonstrate disease progression within one year of treatment initiation.
Another target cellular receptor is type 1 insulin-like growth factor-1 receptor (IGF-1R), IGF-1R is a receptor-tyrosine kinase that plays a critical role in signaling cell survival and proliferation. The IGF system is frequently deregulated in cancer cells by the establishment of autocrine loops involving IGF-I or -II and/or IGF-1R overexpression. Moreover, epidemiological studies have suggested a link between elevated IGF levels and the development of major human cancers, such as breast, colon, lung and prostate cancer. Expression of IGFs and their cognate receptors has been correlated with disease stage, reduced survival, development of metastases and tumor de-differentiation.
Besides IGF-1R, epidermal growth factor receptor (EGFR) has also been implicated in the tumorigenesis of numerous cancers. Effective tumor inhibition has been achieved both experimentally and clinically with a number of strategies that antagonize either receptor activity. Because of the redundancy of growth signaling pathways in tumor cells, inhibition of one receptor function (e.g. EGFR) could be effectively compensated by up-regulation of other growth factor receptor (e.g. IGF-1R)-mediated pathways. For example, a recent study has shown that malignant glioma cell lines expressing equivalent EGFR had significantly different sensitivity to EGFR inhibition depending on their capability of activating IGF-1R and its downstream signaling pathways. Other studies have also demonstrated that overexpression and/or activation of IGF-1R in tumor cells might contribute to their resistance to chemotherapeutic agents, radiation, or antibody therapy such as Herceptin. And consequently, inhibition of IGF-1R signaling has resulted in increased sensitivity of tumor cells to Herceptin.
EGFR is a receptor tyrosine kinase that is expressed on many normal tissues as well as neoplastic lesions of most organs. Overexpression of EGFR or expression of mutant forms of EGFR has been observed in many tumors, particularly epithelial tumors, and is associated with poor clinical prognosis. Inhibition of signaling through this receptor induces an anti-tumor effect. With the FDA approval of Cetuximab, also known as Erbitux (a mouse/human chimeric antibody) in February of 2004, EGFR became an approved antibody drug target for the treatment of metastatic colorectal cancer. In March of 2006, Erbitux also received FDA approval for the treatment squamous cell carcinoma of the head and neck (SCCHN). More recently, Vectibix, a fully human antibody directed against EGFR, was approved for metastatic colorectal cancer. Neither drug is a stand-alone agent in colorectal cancer—they were approved as add-ons to existing colorectal regimens. In colorectal cancer, Erbitux is given in combination with the drug irinotecan and Vectibix is administered after disease progression on, or following fluoropyrimidine-, oxaliplatin-, and irinotecan-containing chemotherapy regimens. Erbitux has been approved as a single agent in recurrent or metastatic SCCHN only where prior platinum-based chemotherapy has failed. Advanced clinical trials which use these drugs to target non-small cell lung carcinoma are ongoing. The sequence of Erbitux or the EGFR antibody, is well known in the art (see for example, Goldstein, et al., Clin. Cancer Res. 1:1311, 1995; U.S. Pat. No. 6,217,866), herein incorporated by reference.
An obstacle in the utilization of a catalytic antibody for selective prodrug activation in cancer therapy has been systemic tumor targeting. The present invention describes an approach based on the adaptation of target binding peptides, or modular recognition domains (MRDs), which are fused to full length antibodies that effectively target tumor cells or soluble molecules while retaining the prodrug activation capability of the catalytic antibody. Since the MRDs are fused to the antibody so as not to significantly mitigate binding to the antibody's traditional binding site, the antibody's specificity remains intact after MRD addition.
As noted in FIG. 2, MRDs, designated by triangles, circles, and squares, can be appended on any of the termini of either heavy or light chains of a typical antibody. The first schematic represents a simple peptibody with a peptide fused to the C-terminus of an Fc. This approach provided for the preparation of bi-, tri-, tetra, and penta-specific antibodies. Display of a single MRD at each N- and C-termini of an IgG provides for octavalent display of the MRD. As an alternative to the construction of bi- and multifunctional antibodies through the combination of antibody variable domains, high-affinity peptides selected from phage display libraries or derived from natural ligands may offer a highly versatile and modular approach to the construction of multifunctional antibodies that retain both the binding and half-life advantages of traditional antibodies. MRDs can also extend the binding capacity of non-catalytic antibodies, providing for an effective approach to extend the binding functionality of antibodies, particularly for therapeutic purposes.